Generating Repetitions with Appropriate Repeated Words

TL;DR

This paper introduces a novel neural approach for generating repetitions in dialogue, utilizing Weighted Label Smoothing and a repetition scoring method to improve the appropriateness of repeated words, validated through automatic and human evaluations.

Contribution

It presents the first neural method for repetition generation in dialogue, incorporating Weighted Label Smoothing and a new scoring technique for better repetition quality.

Findings

Our methods outperform baselines in automatic evaluations.

Human evaluations confirm improved repetition appropriateness.

The approach enhances dialogue trust-building through better repetition generation.

Abstract

A repetition is a response that repeats words in the previous speaker's utterance in a dialogue. Repetitions are essential in communication to build trust with others, as investigated in linguistic studies. In this work, we focus on repetition generation. To the best of our knowledge, this is the first neural approach to address repetition generation. We propose Weighted Label Smoothing, a smoothing method for explicitly learning which words to repeat during fine-tuning, and a repetition scoring method that can output more appropriate repetitions during decoding. We conducted automatic and human evaluations involving applying these methods to the pre-trained language model T5 for generating repetitions. The experimental results indicate that our methods outperformed baselines in both evaluations.